ftrace - Function Tracer ======================== Copyright 2008 Red Hat Inc. Author: Steven Rostedt License: The GNU Free Documentation License, Version 1.2 (dual licensed under the GPL v2) Reviewers: Elias Oltmanns, Randy Dunlap, Andrew Morton, John Kacur, and David Teigland. Written for: 2.6.28-rc2 Introduction ------------ Ftrace is an internal tracer designed to help out developers and designers of systems to find what is going on inside the kernel. It can be used for debugging or analyzing latencies and performance issues that take place outside of user-space. Although ftrace is the function tracer, it also includes an infrastructure that allows for other types of tracing. Some of the tracers that are currently in ftrace include a tracer to trace context switches, the time it takes for a high priority task to run after it was woken up, the time interrupts are disabled, and more (ftrace allows for tracer plugins, which means that the list of tracers can always grow). The File System --------------- Ftrace uses the debugfs file system to hold the control files as well as the files to display output. To mount the debugfs system: # mkdir /debug # mount -t debugfs nodev /debug (Note: it is more common to mount at /sys/kernel/debug, but for simplicity this document will use /debug) That's it! (assuming that you have ftrace configured into your kernel) After mounting the debugfs, you can see a directory called "tracing". This directory contains the control and output files of ftrace. Here is a list of some of the key files: Note: all time values are in microseconds. current_tracer: This is used to set or display the current tracer that is configured. available_tracers: This holds the different types of tracers that have been compiled into the kernel. The tracers listed here can be configured by echoing their name into current_tracer. tracing_enabled: This sets or displays whether the current_tracer is activated and tracing or not. Echo 0 into this file to disable the tracer or 1 to enable it. trace: This file holds the output of the trace in a human readable format (described below). latency_trace: This file shows the same trace but the information is organized more to display possible latencies in the system (described below). trace_pipe: The output is the same as the "trace" file but this file is meant to be streamed with live tracing. Reads from this file will block until new data is retrieved. Unlike the "trace" and "latency_trace" files, this file is a consumer. This means reading from this file causes sequential reads to display more current data. Once data is read from this file, it is consumed, and will not be read again with a sequential read. The "trace" and "latency_trace" files are static, and if the tracer is not adding more data, they will display the same information every time they are read. trace_options: This file lets the user control the amount of data that is displayed in one of the above output files. trace_max_latency: Some of the tracers record the max latency. For example, the time interrupts are disabled. This time is saved in this file. The max trace will also be stored, and displayed by either "trace" or "latency_trace". A new max trace will only be recorded if the latency is greater than the value in this file. (in microseconds) buffer_size_kb: This sets or displays the number of kilobytes each CPU buffer can hold. The tracer buffers are the same size for each CPU. The displayed number is the size of the CPU buffer and not total size of all buffers. The trace buffers are allocated in pages (blocks of memory that the kernel uses for allocation, usually 4 KB in size). If the last page allocated has room for more bytes than requested, the rest of the page will be used, making the actual allocation bigger than requested. (Note, the size may not be a multiple of the page size due to buffer managment overhead.) This can only be updated when the current_tracer is set to "nop". tracing_cpumask: This is a mask that lets the user only trace on specified CPUS. The format is a hex string representing the CPUS. set_ftrace_filter: When dynamic ftrace is configured in (see the section below "dynamic ftrace"), the code is dynamically modified (code text rewrite) to disable calling of the function profiler (mcount). This lets tracing be configured in with practically no overhead in performance. This also has a side effect of enabling or disabling specific functions to be traced. Echoing names of functions into this file will limit the trace to only those functions. set_ftrace_notrace: This has an effect opposite to that of set_ftrace_filter. Any function that is added here will not be traced. If a function exists in both set_ftrace_filter and set_ftrace_notrace, the function will _not_ be traced. available_filter_functions: This lists the functions that ftrace has processed and can trace. These are the function names that you can pass to "set_ftrace_filter" or "set_ftrace_notrace". (See the section "dynamic ftrace" below for more details.) The Tracers ----------- Here is the list of current tracers that may be configured. function - function tracer that uses mcount to trace all functions. sched_switch - traces the context switches between tasks. irqsoff - traces the areas that disable interrupts and saves the trace with the longest max latency. See tracing_max_latency. When a new max is recorded, it replaces the old trace. It is best to view this trace via the latency_trace file. preemptoff - Similar to irqsoff but traces and records the amount of time for which preemption is disabled. preemptirqsoff - Similar to irqsoff and preemptoff, but traces and records the largest time for which irqs and/or preemption is disabled. wakeup - Traces and records the max latency that it takes for the highest priority task to get scheduled after it has been woken up. nop - This is not a tracer. To remove all tracers from tracing simply echo "nop" into current_tracer. Examples of using the tracer ---------------------------- Here are typical examples of using the tracers when controlling them only with the debugfs interface (without using any user-land utilities). Output format: -------------- Here is an example of the output format of the file "trace" -------- # tracer: function # # TASK-PID CPU# TIMESTAMP FUNCTION # | | | | | bash-4251 [01] 10152.583854: path_put <-path_walk bash-4251 [01] 10152.583855: dput <-path_put bash-4251 [01] 10152.583855: _atomic_dec_and_lock <-dput -------- A header is printed with the tracer name that is represented by the trace. In this case the tracer is "function". Then a header showing the format. Task name "bash", the task PID "4251", the CPU that it was running on "01", the timestamp in . format, the function name that was traced "path_put" and the parent function that called this function "path_walk". The timestamp is the time at which the function was entered. The sched_switch tracer also includes tracing of task wakeups and context switches. ksoftirqd/1-7 [01] 1453.070013: 7:115:R + 2916:115:S ksoftirqd/1-7 [01] 1453.070013: 7:115:R + 10:115:S ksoftirqd/1-7 [01] 1453.070013: 7:115:R ==> 10:115:R events/1-10 [01] 1453.070013: 10:115:S ==> 2916:115:R kondemand/1-2916 [01] 1453.070013: 2916:115:S ==> 7:115:R ksoftirqd/1-7 [01] 1453.070013: 7:115:S ==> 0:140:R Wake ups are represented by a "+" and the context switches are shown as "==>". The format is: Context switches: Previous task Next Task :: ==> :: Wake ups: Current task Task waking up :: + :: The prio is the internal kernel priority, which is the inverse of the priority that is usually displayed by user-space tools. Zero represents the highest priority (99). Prio 100 starts the "nice" priorities with 100 being equal to nice -20 and 139 being nice 19. The prio "140" is reserved for the idle task which is the lowest priority thread (pid 0). Latency trace format -------------------- For traces that display latency times, the latency_trace file gives somewhat more information to see why a latency happened. Here is a typical trace. # tracer: irqsoff # irqsoff latency trace v1.1.5 on 2.6.26-rc8 -------------------------------------------------------------------- latency: 97 us, #3/3, CPU#0 | (M:preempt VP:0, KP:0, SP:0 HP:0 #P:2) ----------------- | task: swapper-0 (uid:0 nice:0 policy:0 rt_prio:0) ----------------- => started at: apic_timer_interrupt => ended at: do_softirq # _------=> CPU# # / _-----=> irqs-off # | / _----=> need-resched # || / _---=> hardirq/softirq # ||| / _--=> preempt-depth # |||| / # ||||| delay # cmd pid ||||| time | caller # \ / ||||| \ | / -0 0d..1 0us+: trace_hardirqs_off_thunk (apic_timer_interrupt) -0 0d.s. 97us : __do_softirq (do_softirq) -0 0d.s1 98us : trace_hardirqs_on (do_softirq) This shows that the current tracer is "irqsoff" tracing the time for which interrupts were disabled. It gives the trace version and the version of the kernel upon which this was executed on (2.6.26-rc8). Then it displays the max latency in microsecs (97 us). The number of trace entries displayed and the total number recorded (both are three: #3/3). The type of preemption that was used (PREEMPT). VP, KP, SP, and HP are always zero and are reserved for later use. #P is the number of online CPUS (#P:2). The task is the process that was running when the latency occurred. (swapper pid: 0). The start and stop (the functions in which the interrupts were disabled and enabled respectively) that caused the latencies: apic_timer_interrupt is where the interrupts were disabled. do_softirq is where they were enabled again. The next lines after the header are the trace itself. The header explains which is which. cmd: The name of the process in the trace. pid: The PID of that process. CPU#: The CPU which the process was running on. irqs-off: 'd' interrupts are disabled. '.' otherwise. Note: If the architecture does not support a way to read the irq flags variable, an 'X' will always be printed here. need-resched: 'N' task need_resched is set, '.' otherwise. hardirq/softirq: 'H' - hard irq occurred inside a softirq. 'h' - hard irq is running 's' - soft irq is running '.' - normal context. preempt-depth: The level of preempt_disabled The above is mostly meaningful for kernel developers. time: This differs from the trace file output. The trace file output includes an absolute timestamp. The timestamp used by the latency_trace file is relative to the start of the trace. delay: This is just to help catch your eye a bit better. And needs to be fixed to be only relative to the same CPU. The marks are determined by the difference between this current trace and the next trace. '!' - greater than preempt_mark_thresh (default 100) '+' - greater than 1 microsecond ' ' - less than or equal to 1 microsecond. The rest is the same as the 'trace' file. trace_options ------------- The trace_options file is used to control what gets printed in the trace output. To see what is available, simply cat the file: cat /debug/tracing/trace_options print-parent nosym-offset nosym-addr noverbose noraw nohex nobin \ noblock nostacktrace nosched-tree nouserstacktrace nosym-userobj To disable one of the options, echo in the option prepended with "no". echo noprint-parent > /debug/tracing/trace_options To enable an option, leave off the "no". echo sym-offset > /debug/tracing/trace_options Here are the available options: print-parent - On function traces, display the calling function as well as the function being traced. print-parent: bash-4000 [01] 1477.606694: simple_strtoul <-strict_strtoul noprint-parent: bash-4000 [01] 1477.606694: simple_strtoul sym-offset - Display not only the function name, but also the offset in the function. For example, instead of seeing just "ktime_get", you will see "ktime_get+0xb/0x20". sym-offset: bash-4000 [01] 1477.606694: simple_strtoul+0x6/0xa0 sym-addr - this will also display the function address as well as the function name. sym-addr: bash-4000 [01] 1477.606694: simple_strtoul verbose - This deals with the latency_trace file. bash 4000 1 0 00000000 00010a95 [58127d26] 1720.415ms \ (+0.000ms): simple_strtoul (strict_strtoul) raw - This will display raw numbers. This option is best for use with user applications that can translate the raw numbers better than having it done in the kernel. hex - Similar to raw, but the numbers will be in a hexadecimal format. bin - This will print out the formats in raw binary. block - TBD (needs update) stacktrace - This is one of the options that changes the trace itself. When a trace is recorded, so is the stack of functions. This allows for back traces of trace sites. userstacktrace - This option changes the trace. It records a stacktrace of the current userspace thread. sym-userobj - when user stacktrace are enabled, look up which object the address belongs to, and print a relative address This is especially useful when ASLR is on, otherwise you don't get a chance to resolve the address to object/file/line after the app is no longer running The lookup is performed when you read trace,trace_pipe,latency_trace. Example: a.out-1623 [000] 40874.465068: /root/a.out[+0x480] <-/root/a.out[+0 x494] <- /root/a.out[+0x4a8] <- /lib/libc-2.7.so[+0x1e1a6] sched-tree - TBD (any users??) sched_switch ------------ This tracer simply records schedule switches. Here is an example of how to use it. # echo sched_switch > /debug/tracing/current_tracer # echo 1 > /debug/tracing/tracing_enabled # sleep 1 # echo 0 > /debug/tracing/tracing_enabled # cat /debug/tracing/trace # tracer: sched_switch # # TASK-PID CPU# TIMESTAMP FUNCTION # | | | | | bash-3997 [01] 240.132281: 3997:120:R + 4055:120:R bash-3997 [01] 240.132284: 3997:120:R ==> 4055:120:R sleep-4055 [01] 240.132371: 4055:120:S ==> 3997:120:R bash-3997 [01] 240.132454: 3997:120:R + 4055:120:S bash-3997 [01] 240.132457: 3997:120:R ==> 4055:120:R sleep-4055 [01] 240.132460: 4055:120:D ==> 3997:120:R bash-3997 [01] 240.132463: 3997:120:R + 4055:120:D bash-3997 [01] 240.132465: 3997:120:R ==> 4055:120:R -0 [00] 240.132589: 0:140:R + 4:115:S -0 [00] 240.132591: 0:140:R ==> 4:115:R ksoftirqd/0-4 [00] 240.132595: 4:115:S ==> 0:140:R -0 [00] 240.132598: 0:140:R + 4:115:S -0 [00] 240.132599: 0:140:R ==> 4:115:R ksoftirqd/0-4 [00] 240.132603: 4:115:S ==> 0:140:R sleep-4055 [01] 240.133058: 4055:120:S ==> 3997:120:R [...] As we have discussed previously about this format, the header shows the name of the trace and points to the options. The "FUNCTION" is a misnomer since here it represents the wake ups and context switches. The sched_switch file only lists the wake ups (represented with '+') and context switches ('==>') with the previous task or current task first followed by the next task or task waking up. The format for both of these is PID:KERNEL-PRIO:TASK-STATE. Remember that the KERNEL-PRIO is the inverse of the actual priority with zero (0) being the highest priority and the nice values starting at 100 (nice -20). Below is a quick chart to map the kernel priority to user land priorities. Kernel priority: 0 to 99 ==> user RT priority 99 to 0 Kernel priority: 100 to 139 ==> user nice -20 to 19 Kernel priority: 140 ==> idle task priority The task states are: R - running : wants to run, may not actually be running S - sleep : process is waiting to be woken up (handles signals) D - disk sleep (uninterruptible sleep) : process must be woken up (ignores signals) T - stopped : process suspended t - traced : process is being traced (with something like gdb) Z - zombie : process waiting to be cleaned up X - unknown ftrace_enabled -------------- The following tracers (listed below) give different output depending on whether or not the sysctl ftrace_enabled is set. To set ftrace_enabled, one can either use the sysctl function or set it via the proc file system interface. sysctl kernel.ftrace_enabled=1 or echo 1 > /proc/sys/kernel/ftrace_enabled To disable ftrace_enabled simply replace the '1' with '0' in the above commands. When ftrace_enabled is set the tracers will also record the functions that are within the trace. The descriptions of the tracers will also show an example with ftrace enabled. irqsoff ------- When interrupts are disabled, the CPU can not react to any other external event (besides NMIs and SMIs). This prevents the timer interrupt from triggering or the mouse interrupt from letting the kernel know of a new mouse event. The result is a latency with the reaction time. The irqsoff tracer tracks the time for which interrupts are disabled. When a new maximum latency is hit, the tracer saves the trace leading up to that latency point so that every time a new maximum is reached, the old saved trace is discarded and the new trace is saved. To reset the maximum, echo 0 into tracing_max_latency. Here is an example: # echo irqsoff > /debug/tracing/current_tracer # echo 0 > /debug/tracing/tracing_max_latency # echo 1 > /debug/tracing/tracing_enabled # ls -ltr [...] # echo 0 > /debug/tracing/tracing_enabled # cat /debug/tracing/latency_trace # tracer: irqsoff # irqsoff latency trace v1.1.5 on 2.6.26 -------------------------------------------------------------------- latency: 12 us, #3/3, CPU#1 | (M:preempt VP:0, KP:0, SP:0 HP:0 #P:2) ----------------- | task: bash-3730 (uid:0 nice:0 policy:0 rt_prio:0) ----------------- => started at: sys_setpgid => ended at: sys_setpgid # _------=> CPU# # / _-----=> irqs-off # | / _----=> need-resched # || / _---=> hardirq/softirq # ||| / _--=> preempt-depth # |||| / # ||||| delay # cmd pid ||||| time | caller # \ / ||||| \ | / bash-3730 1d... 0us : _write_lock_irq (sys_setpgid) bash-3730 1d..1 1us+: _write_unlock_irq (sys_setpgid) bash-3730 1d..2 14us : trace_hardirqs_on (sys_setpgid) Here we see that that we had a latency of 12 microsecs (which is very good). The _write_lock_irq in sys_setpgid disabled interrupts. The difference between the 12 and the displayed timestamp 14us occurred because the clock was incremented between the time of recording the max latency and the time of recording the function that had that latency. Note the above example had ftrace_enabled not set. If we set the ftrace_enabled, we get a much larger output: # tracer: irqsoff # irqsoff latency trace v1.1.5 on 2.6.26-rc8 -------------------------------------------------------------------- latency: 50 us, #101/101, CPU#0 | (M:preempt VP:0, KP:0, SP:0 HP:0 #P:2) ----------------- | task: ls-4339 (uid:0 nice:0 policy:0 rt_prio:0) ----------------- => started at: __alloc_pages_internal => ended at: __alloc_pages_internal # _------=> CPU# # / _-----=> irqs-off # | / _----=> need-resched # || / _---=> hardirq/softirq # ||| / _--=> preempt-depth # |||| / # ||||| delay # cmd pid ||||| time | caller # \ / ||||| \ | / ls-4339 0...1 0us+: get_page_from_freelist (__alloc_pages_internal) ls-4339 0d..1 3us : rmqueue_bulk (get_page_from_freelist) ls-4339 0d..1 3us : _spin_lock (rmqueue_bulk) ls-4339 0d..1 4us : add_preempt_count (_spin_lock) ls-4339 0d..2 4us : __rmqueue (rmqueue_bulk) ls-4339 0d..2 5us : __rmqueue_smallest (__rmqueue) ls-4339 0d..2 5us : __mod_zone_page_state (__rmqueue_smallest) ls-4339 0d..2 6us : __rmqueue (rmqueue_bulk) ls-4339 0d..2 6us : __rmqueue_smallest (__rmqueue) ls-4339 0d..2 7us : __mod_zone_page_state (__rmqueue_smallest) ls-4339 0d..2 7us : __rmqueue (rmqueue_bulk) ls-4339 0d..2 8us : __rmqueue_smallest (__rmqueue) [...] ls-4339 0d..2 46us : __rmqueue_smallest (__rmqueue) ls-4339 0d..2 47us : __mod_zone_page_state (__rmqueue_smallest) ls-4339 0d..2 47us : __rmqueue (rmqueue_bulk) ls-4339 0d..2 48us : __rmqueue_smallest (__rmqueue) ls-4339 0d..2 48us : __mod_zone_page_state (__rmqueue_smallest) ls-4339 0d..2 49us : _spin_unlock (rmqueue_bulk) ls-4339 0d..2 49us : sub_preempt_count (_spin_unlock) ls-4339 0d..1 50us : get_page_from_freelist (__alloc_pages_internal) ls-4339 0d..2 51us : trace_hardirqs_on (__alloc_pages_internal) Here we traced a 50 microsecond latency. But we also see all the functions that were called during that time. Note that by enabling function tracing, we incur an added overhead. This overhead may extend the latency times. But nevertheless, this trace has provided some very helpful debugging information. preemptoff ---------- When preemption is disabled, we may be able to receive interrupts but the task cannot be preempted and a higher priority task must wait for preemption to be enabled again before it can preempt a lower priority task. The preemptoff tracer traces the places that disable preemption. Like the irqsoff tracer, it records the maximum latency for which preemption was disabled. The control of preemptoff tracer is much like the irqsoff tracer. # echo preemptoff > /debug/tracing/current_tracer # echo 0 > /debug/tracing/tracing_max_latency # echo 1 > /debug/tracing/tracing_enabled # ls -ltr [...] # echo 0 > /debug/tracing/tracing_enabled # cat /debug/tracing/latency_trace # tracer: preemptoff # preemptoff latency trace v1.1.5 on 2.6.26-rc8 -------------------------------------------------------------------- latency: 29 us, #3/3, CPU#0 | (M:preempt VP:0, KP:0, SP:0 HP:0 #P:2) ----------------- | task: sshd-4261 (uid:0 nice:0 policy:0 rt_prio:0) ----------------- => started at: do_IRQ => ended at: __do_softirq # _------=> CPU# # / _-----=> irqs-off # | / _----=> need-resched # || / _---=> hardirq/softirq # ||| / _--=> preempt-depth # |||| / # ||||| delay # cmd pid ||||| time | caller # \ / ||||| \ | / sshd-4261 0d.h. 0us+: irq_enter (do_IRQ) sshd-4261 0d.s. 29us : _local_bh_enable (__do_softirq) sshd-4261 0d.s1 30us : trace_preempt_on (__do_softirq) This has some more changes. Preemption was disabled when an interrupt came in (notice the 'h'), and was enabled while doing a softirq. (notice the 's'). But we also see that interrupts have been disabled when entering the preempt off section and leaving it (the 'd'). We do not know if interrupts were enabled in the mean time. # tracer: preemptoff # preemptoff latency trace v1.1.5 on 2.6.26-rc8 -------------------------------------------------------------------- latency: 63 us, #87/87, CPU#0 | (M:preempt VP:0, KP:0, SP:0 HP:0 #P:2) ----------------- | task: sshd-4261 (uid:0 nice:0 policy:0 rt_prio:0) ----------------- => started at: remove_wait_queue => ended at: __do_softirq # _------=> CPU# # / _-----=> irqs-off # | / _----=> need-resched # || / _---=> hardirq/softirq # ||| / _--=> preempt-depth # |||| / # ||||| delay # cmd pid ||||| time | caller # \ / ||||| \ | / sshd-4261 0d..1 0us : _spin_lock_irqsave (remove_wait_queue) sshd-4261 0d..1 1us : _spin_unlock_irqrestore (remove_wait_queue) sshd-4261 0d..1 2us : do_IRQ (common_interrupt) sshd-4261 0d..1 2us : irq_enter (do_IRQ) sshd-4261 0d..1 2us : idle_cpu (irq_enter) sshd-4261 0d..1 3us : add_preempt_count (irq_enter) sshd-4261 0d.h1 3us : idle_cpu (irq_enter) sshd-4261 0d.h. 4us : handle_fasteoi_irq (do_IRQ) [...] sshd-4261 0d.h. 12us : add_preempt_count (_spin_lock) sshd-4261 0d.h1 12us : ack_ioapic_quirk_irq (handle_fasteoi_irq) sshd-4261 0d.h1 13us : move_native_irq (ack_ioapic_quirk_irq) sshd-4261 0d.h1 13us : _spin_unlock (handle_fasteoi_irq) sshd-4261 0d.h1 14us : sub_preempt_count (_spin_unlock) sshd-4261 0d.h1 14us : irq_exit (do_IRQ) sshd-4261 0d.h1 15us : sub_preempt_count (irq_exit) sshd-4261 0d..2 15us : do_softirq (irq_exit) sshd-4261 0d... 15us : __do_softirq (do_softirq) sshd-4261 0d... 16us : __local_bh_disable (__do_softirq) sshd-4261 0d... 16us+: add_preempt_count (__local_bh_disable) sshd-4261 0d.s4 20us : add_preempt_count (__local_bh_disable) sshd-4261 0d.s4 21us : sub_preempt_count (local_bh_enable) sshd-4261 0d.s5 21us : sub_preempt_count (local_bh_enable) [...] sshd-4261 0d.s6 41us : add_preempt_count (__local_bh_disable) sshd-4261 0d.s6 42us : sub_preempt_count (local_bh_enable) sshd-4261 0d.s7 42us : sub_preempt_count (local_bh_enable) sshd-4261 0d.s5 43us : add_preempt_count (__local_bh_disable) sshd-4261 0d.s5 43us : sub_preempt_count (local_bh_enable_ip) sshd-4261 0d.s6 44us : sub_preempt_count (local_bh_enable_ip) sshd-4261 0d.s5 44us : add_preempt_count (__local_bh_disable) sshd-4261 0d.s5 45us : sub_preempt_count (local_bh_enable) [...] sshd-4261 0d.s. 63us : _local_bh_enable (__do_softirq) sshd-4261 0d.s1 64us : trace_preempt_on (__do_softirq) The above is an example of the preemptoff trace with ftrace_enabled set. Here we see that interrupts were disabled the entire time. The irq_enter code lets us know that we entered an interrupt 'h'. Before that, the functions being traced still show that it is not in an interrupt, but we can see from the functions themselves that this is not the case. Notice that __do_softirq when called does not have a preempt_count. It may seem that we missed a preempt enabling. What really happened is that the preempt count is held on the thread's stack and we switched to the softirq stack (4K stacks in effect). The code does not copy the preempt count, but because interrupts are disabled, we do not need to worry about it. Having a tracer like this is good for letting people know what really happens inside the kernel. preemptirqsoff -------------- Knowing the locations that have interrupts disabled or preemption disabled for the longest times is helpful. But sometimes we would like to know when either preemption and/or interrupts are disabled. Consider the following code: local_irq_disable(); call_function_with_irqs_off(); preempt_disable(); call_function_with_irqs_and_preemption_off(); local_irq_enable(); call_function_with_preemption_off(); preempt_enable(); The irqsoff tracer will record the total length of call_function_with_irqs_off() and call_function_with_irqs_and_preemption_off(). The preemptoff tracer will record the total length of call_function_with_irqs_and_preemption_off() and call_function_with_preemption_off(). But neither will trace the time that interrupts and/or preemption is disabled. This total time is the time that we can not schedule. To record this time, use the preemptirqsoff tracer. Again, using this trace is much like the irqsoff and preemptoff tracers. # echo preemptirqsoff > /debug/tracing/current_tracer # echo 0 > /debug/tracing/tracing_max_latency # echo 1 > /debug/tracing/tracing_enabled # ls -ltr [...] # echo 0 > /debug/tracing/tracing_enabled # cat /debug/tracing/latency_trace # tracer: preemptirqsoff # preemptirqsoff latency trace v1.1.5 on 2.6.26-rc8 -------------------------------------------------------------------- latency: 293 us, #3/3, CPU#0 | (M:preempt VP:0, KP:0, SP:0 HP:0 #P:2) ----------------- | task: ls-4860 (uid:0 nice:0 policy:0 rt_prio:0) ----------------- => started at: apic_timer_interrupt => ended at: __do_softirq # _------=> CPU# # / _-----=> irqs-off # | / _----=> need-resched # || / _---=> hardirq/softirq # ||| / _--=> preempt-depth # |||| / # ||||| delay # cmd pid ||||| time | caller # \ / ||||| \ | / ls-4860 0d... 0us!: trace_hardirqs_off_thunk (apic_timer_interrupt) ls-4860 0d.s. 294us : _local_bh_enable (__do_softirq) ls-4860 0d.s1 294us : trace_preempt_on (__do_softirq) The trace_hardirqs_off_thunk is called from assembly on x86 when interrupts are disabled in the assembly code. Without the function tracing, we do not know if interrupts were enabled within the preemption points. We do see that it started with preemption enabled. Here is a trace with ftrace_enabled set: # tracer: preemptirqsoff # preemptirqsoff latency trace v1.1.5 on 2.6.26-rc8 -------------------------------------------------------------------- latency: 105 us, #183/183, CPU#0 | (M:preempt VP:0, KP:0, SP:0 HP:0 #P:2) ----------------- | task: sshd-4261 (uid:0 nice:0 policy:0 rt_prio:0) ----------------- => started at: write_chan => ended at: __do_softirq # _------=> CPU# # / _-----=> irqs-off # | / _----=> need-resched # || / _---=> hardirq/softirq # ||| / _--=> preempt-depth # |||| / # ||||| delay # cmd pid ||||| time | caller # \ / ||||| \ | / ls-4473 0.N.. 0us : preempt_schedule (write_chan) ls-4473 0dN.1 1us : _spin_lock (schedule) ls-4473 0dN.1 2us : add_preempt_count (_spin_lock) ls-4473 0d..2 2us : put_prev_task_fair (schedule) [...] ls-4473 0d..2 13us : set_normalized_timespec (ktime_get_ts) ls-4473 0d..2 13us : __switch_to (schedule) sshd-4261 0d..2 14us : finish_task_switch (schedule) sshd-4261 0d..2 14us : _spin_unlock_irq (finish_task_switch) sshd-4261 0d..1 15us : add_preempt_count (_spin_lock_irqsave) sshd-4261 0d..2 16us : _spin_unlock_irqrestore (hrtick_set) sshd-4261 0d..2 16us : do_IRQ (common_interrupt) sshd-4261 0d..2 17us : irq_enter (do_IRQ) sshd-4261 0d..2 17us : idle_cpu (irq_enter) sshd-4261 0d..2 18us : add_preempt_count (irq_enter) sshd-4261 0d.h2 18us : idle_cpu (irq_enter) sshd-4261 0d.h. 18us : handle_fasteoi_irq (do_IRQ) sshd-4261 0d.h. 19us : _spin_lock (handle_fasteoi_irq) sshd-4261 0d.h. 19us : add_preempt_count (_spin_lock) sshd-4261 0d.h1 20us : _spin_unlock (handle_fasteoi_irq) sshd-4261 0d.h1 20us : sub_preempt_count (_spin_unlock) [...] sshd-4261 0d.h1 28us : _spin_unlock (handle_fasteoi_irq) sshd-4261 0d.h1 29us : sub_preempt_count (_spin_unlock) sshd-4261 0d.h2 29us : irq_exit (do_IRQ) sshd-4261 0d.h2 29us : sub_preempt_count (irq_exit) sshd-4261 0d..3 30us : do_softirq (irq_exit) sshd-4261 0d... 30us : __do_softirq (do_softirq) sshd-4261 0d... 31us : __local_bh_disable (__do_softirq) sshd-4261 0d... 31us+: add_preempt_count (__local_bh_disable) sshd-4261 0d.s4 34us : add_preempt_count (__local_bh_disable) [...] sshd-4261 0d.s3 43us : sub_preempt_count (local_bh_enable_ip) sshd-4261 0d.s4 44us : sub_preempt_count (local_bh_enable_ip) sshd-4261 0d.s3 44us : smp_apic_timer_interrupt (apic_timer_interrupt) sshd-4261 0d.s3 45us : irq_enter (smp_apic_timer_interrupt) sshd-4261 0d.s3 45us : idle_cpu (irq_enter) sshd-4261 0d.s3 46us : add_preempt_count (irq_enter) sshd-4261 0d.H3 46us : idle_cpu (irq_enter) sshd-4261 0d.H3 47us : hrtimer_interrupt (smp_apic_timer_interrupt) sshd-4261 0d.H3 47us : ktime_get (hrtimer_interrupt) [...] sshd-4261 0d.H3 81us : tick_program_event (hrtimer_interrupt) sshd-4261 0d.H3 82us : ktime_get (tick_program_event) sshd-4261 0d.H3 82us : ktime_get_ts (ktime_get) sshd-4261 0d.H3 83us : getnstimeofday (ktime_get_ts) sshd-4261 0d.H3 83us : set_normalized_timespec (ktime_get_ts) sshd-4261 0d.H3 84us : clockevents_program_event (tick_program_event) sshd-4261 0d.H3 84us : lapic_next_event (clockevents_program_event) sshd-4261 0d.H3 85us : irq_exit (smp_apic_timer_interrupt) sshd-4261 0d.H3 85us : sub_preempt_count (irq_exit) sshd-4261 0d.s4 86us : sub_preempt_count (irq_exit) sshd-4261 0d.s3 86us : add_preempt_count (__local_bh_disable) [...] sshd-4261 0d.s1 98us : sub_preempt_count (net_rx_action) sshd-4261 0d.s. 99us : add_preempt_count (_spin_lock_irq) sshd-4261 0d.s1 99us+: _spin_unlock_irq (run_timer_softirq) sshd-4261 0d.s. 104us : _local_bh_enable (__do_softirq) sshd-4261 0d.s. 104us : sub_preempt_count (_local_bh_enable) sshd-4261 0d.s. 105us : _local_bh_enable (__do_softirq) sshd-4261 0d.s1 105us : trace_preempt_on (__do_softirq) This is a very interesting trace. It started with the preemption of the ls task. We see that the task had the "need_resched" bit set via the 'N' in the trace. Interrupts were disabled before the spin_lock at the beginning of the trace. We see that a schedule took place to run sshd. When the interrupts were enabled, we took an interrupt. On return from the interrupt handler, the softirq ran. We took another interrupt while running the softirq as we see from the capital 'H'. wakeup ------ In a Real-Time environment it is very important to know the wakeup time it takes for the highest priority task that is woken up to the time that it executes. This is also known as "schedule latency". I stress the point that this is about RT tasks. It is also important to know the scheduling latency of non-RT tasks, but the average schedule latency is better for non-RT tasks. Tools like LatencyTop are more appropriate for such measurements. Real-Time environments are interested in the worst case latency. That is the longest latency it takes for something to happen, and not the average. We can have a very fast scheduler that may only have a large latency once in a while, but that would not work well with Real-Time tasks. The wakeup tracer was designed to record the worst case wakeups of RT tasks. Non-RT tasks are not recorded because the tracer only records one worst case and tracing non-RT tasks that are unpredictable will overwrite the worst case latency of RT tasks. Since this tracer only deals with RT tasks, we will run this slightly differently than we did with the previous tracers. Instead of performing an 'ls', we will run 'sleep 1' under 'chrt' which changes the priority of the task. # echo wakeup > /debug/tracing/current_tracer # echo 0 > /debug/tracing/tracing_max_latency # echo 1 > /debug/tracing/tracing_enabled # chrt -f 5 sleep 1 # echo 0 > /debug/tracing/tracing_enabled # cat /debug/tracing/latency_trace # tracer: wakeup # wakeup latency trace v1.1.5 on 2.6.26-rc8 -------------------------------------------------------------------- latency: 4 us, #2/2, CPU#1 | (M:preempt VP:0, KP:0, SP:0 HP:0 #P:2) ----------------- | task: sleep-4901 (uid:0 nice:0 policy:1 rt_prio:5) ----------------- # _------=> CPU# # / _-----=> irqs-off # | / _----=> need-resched # || / _---=> hardirq/softirq # ||| / _--=> preempt-depth # |||| / # ||||| delay # cmd pid ||||| time | caller # \ / ||||| \ | / -0 1d.h4 0us+: try_to_wake_up (wake_up_process) -0 1d..4 4us : schedule (cpu_idle) Running this on an idle system, we see that it only took 4 microseconds to perform the task switch. Note, since the trace marker in the schedule is before the actual "switch", we stop the tracing when the recorded task is about to schedule in. This may change if we add a new marker at the end of the scheduler. Notice that the recorded task is 'sleep' with the PID of 4901 and it has an rt_prio of 5. This priority is user-space priority and not the internal kernel priority. The policy is 1 for SCHED_FIFO and 2 for SCHED_RR. Doing the same with chrt -r 5 and ftrace_enabled set. # tracer: wakeup # wakeup latency trace v1.1.5 on 2.6.26-rc8 -------------------------------------------------------------------- latency: 50 us, #60/60, CPU#1 | (M:preempt VP:0, KP:0, SP:0 HP:0 #P:2) ----------------- | task: sleep-4068 (uid:0 nice:0 policy:2 rt_prio:5) ----------------- # _------=> CPU# # / _-----=> irqs-off # | / _----=> need-resched # || / _---=> hardirq/softirq # ||| / _--=> preempt-depth # |||| / # ||||| delay # cmd pid ||||| time | caller # \ / ||||| \ | / ksoftirq-7 1d.H3 0us : try_to_wake_up (wake_up_process) ksoftirq-7 1d.H4 1us : sub_preempt_count (marker_probe_cb) ksoftirq-7 1d.H3 2us : check_preempt_wakeup (try_to_wake_up) ksoftirq-7 1d.H3 3us : update_curr (check_preempt_wakeup) ksoftirq-7 1d.H3 4us : calc_delta_mine (update_curr) ksoftirq-7 1d.H3 5us : __resched_task (check_preempt_wakeup) ksoftirq-7 1d.H3 6us : task_wake_up_rt (try_to_wake_up) ksoftirq-7 1d.H3 7us : _spin_unlock_irqrestore (try_to_wake_up) [...] ksoftirq-7 1d.H2 17us : irq_exit (smp_apic_timer_interrupt) ksoftirq-7 1d.H2 18us : sub_preempt_count (irq_exit) ksoftirq-7 1d.s3 19us : sub_preempt_count (irq_exit) ksoftirq-7 1..s2 20us : rcu_process_callbacks (__do_softirq) [...] ksoftirq-7 1..s2 26us : __rcu_process_callbacks (rcu_process_callbacks) ksoftirq-7 1d.s2 27us : _local_bh_enable (__do_softirq) ksoftirq-7 1d.s2 28us : sub_preempt_count (_local_bh_enable) ksoftirq-7 1.N.3 29us : sub_preempt_count (ksoftirqd) ksoftirq-7 1.N.2 30us : _cond_resched (ksoftirqd) ksoftirq-7 1.N.2 31us : __cond_resched (_cond_resched) ksoftirq-7 1.N.2 32us : add_preempt_count (__cond_resched) ksoftirq-7 1.N.2 33us : schedule (__cond_resched) ksoftirq-7 1.N.2 33us : add_preempt_count (schedule) ksoftirq-7 1.N.3 34us : hrtick_clear (schedule) ksoftirq-7 1dN.3 35us : _spin_lock (schedule) ksoftirq-7 1dN.3 36us : add_preempt_count (_spin_lock) ksoftirq-7 1d..4 37us : put_prev_task_fair (schedule) ksoftirq-7 1d..4 38us : update_curr (put_prev_task_fair) [...] ksoftirq-7 1d..5 47us : _spin_trylock (tracing_record_cmdline) ksoftirq-7 1d..5 48us : add_preempt_count (_spin_trylock) ksoftirq-7 1d..6 49us : _spin_unlock (tracing_record_cmdline) ksoftirq-7 1d..6 49us : sub_preempt_count (_spin_unlock) ksoftirq-7 1d..4 50us : schedule (__cond_resched) The interrupt went off while running ksoftirqd. This task runs at SCHED_OTHER. Why did not we see the 'N' set early? This may be a harmless bug with x86_32 and 4K stacks. On x86_32 with 4K stacks configured, the interrupt and softirq run with their own stack. Some information is held on the top of the task's stack (need_resched and preempt_count are both stored there). The setting of the NEED_RESCHED bit is done directly to the task's stack, but the reading of the NEED_RESCHED is done by looking at the current stack, which in this case is the stack for the hard interrupt. This hides the fact that NEED_RESCHED has been set. We do not see the 'N' until we switch back to the task's assigned stack. function -------- This tracer is the function tracer. Enabling the function tracer can be done from the debug file system. Make sure the ftrace_enabled is set; otherwise this tracer is a nop. # sysctl kernel.ftrace_enabled=1 # echo function > /debug/tracing/current_tracer # echo 1 > /debug/tracing/tracing_enabled # usleep 1 # echo 0 > /debug/tracing/tracing_enabled # cat /debug/tracing/trace # tracer: function # # TASK-PID CPU# TIMESTAMP FUNCTION # | | | | | bash-4003 [00] 123.638713: finish_task_switch <-schedule bash-4003 [00] 123.638714: _spin_unlock_irq <-finish_task_switch bash-4003 [00] 123.638714: sub_preempt_count <-_spin_unlock_irq bash-4003 [00] 123.638715: hrtick_set <-schedule bash-4003 [00] 123.638715: _spin_lock_irqsave <-hrtick_set bash-4003 [00] 123.638716: add_preempt_count <-_spin_lock_irqsave bash-4003 [00] 123.638716: _spin_unlock_irqrestore <-hrtick_set bash-4003 [00] 123.638717: sub_preempt_count <-_spin_unlock_irqrestore bash-4003 [00] 123.638717: hrtick_clear <-hrtick_set bash-4003 [00] 123.638718: sub_preempt_count <-schedule bash-4003 [00] 123.638718: sub_preempt_count <-preempt_schedule bash-4003 [00] 123.638719: wait_for_completion <-__stop_machine_run bash-4003 [00] 123.638719: wait_for_common <-wait_for_completion bash-4003 [00] 123.638720: _spin_lock_irq <-wait_for_common bash-4003 [00] 123.638720: add_preempt_count <-_spin_lock_irq [...] Note: function tracer uses ring buffers to store the above entries. The newest data may overwrite the oldest data. Sometimes using echo to stop the trace is not sufficient because the tracing could have overwritten the data that you wanted to record. For this reason, it is sometimes better to disable tracing directly from a program. This allows you to stop the tracing at the point that you hit the part that you are interested in. To disable the tracing directly from a C program, something like following code snippet can be used: int trace_fd; [...] int main(int argc, char *argv[]) { [...] trace_fd = open("/debug/tracing/tracing_enabled", O_WRONLY); [...] if (condition_hit()) { write(trace_fd, "0", 1); } [...] } Note: Here we hard coded the path name. The debugfs mount is not guaranteed to be at /debug (and is more commonly at /sys/kernel/debug). For simple one time traces, the above is sufficent. For anything else, a search through /proc/mounts may be needed to find where the debugfs file-system is mounted. dynamic ftrace -------------- If CONFIG_DYNAMIC_FTRACE is set, the system will run with virtually no overhead when function tracing is disabled. The way this works is the mcount function call (placed at the start of every kernel function, produced by the -pg switch in gcc), starts of pointing to a simple return. (Enabling FTRACE will include the -pg switch in the compiling of the kernel.) At compile time every C file object is run through the recordmcount.pl script (located in the scripts directory). This script will process the C object using objdump to find all the locations in the .text section that call mcount. (Note, only the .text section is processed, since processing other sections like .init.text may cause races due to those sections being freed). A new section called "__mcount_loc" is created that holds references to all the mcount call sites in the .text section. This section is compiled back into the original object. The final linker will add all these references into a single table. On boot up, before SMP is initialized, the dynamic ftrace code scans this table and updates all the locations into nops. It also records the locations, which are added to the available_filter_functions list. Modules are processed as they are loaded and before they are executed. When a module is unloaded, it also removes its functions from the ftrace function list. This is automatic in the module unload code, and the module author does not need to worry about it. When tracing is enabled, kstop_machine is called to prevent races with the CPUS executing code being modified (which can cause the CPU to do undesireable things), and the nops are patched back to calls. But this time, they do not call mcount (which is just a function stub). They now call into the ftrace infrastructure. One special side-effect to the recording of the functions being traced is that we can now selectively choose which functions we wish to trace and which ones we want the mcount calls to remain as nops. Two files are used, one for enabling and one for disabling the tracing of specified functions. They are: set_ftrace_filter and set_ftrace_notrace A list of available functions that you can add to these files is listed in: available_filter_functions # cat /debug/tracing/available_filter_functions put_prev_task_idle kmem_cache_create pick_next_task_rt get_online_cpus pick_next_task_fair mutex_lock [...] If I am only interested in sys_nanosleep and hrtimer_interrupt: # echo sys_nanosleep hrtimer_interrupt \ > /debug/tracing/set_ftrace_filter # echo ftrace > /debug/tracing/current_tracer # echo 1 > /debug/tracing/tracing_enabled # usleep 1 # echo 0 > /debug/tracing/tracing_enabled # cat /debug/tracing/trace # tracer: ftrace # # TASK-PID CPU# TIMESTAMP FUNCTION # | | | | | usleep-4134 [00] 1317.070017: hrtimer_interrupt <-smp_apic_timer_interrupt usleep-4134 [00] 1317.070111: sys_nanosleep <-syscall_call -0 [00] 1317.070115: hrtimer_interrupt <-smp_apic_timer_interrupt To see which functions are being traced, you can cat the file: # cat /debug/tracing/set_ftrace_filter hrtimer_interrupt sys_nanosleep Perhaps this is not enough. The filters also allow simple wild cards. Only the following are currently available * - will match functions that begin with * - will match functions that end with ** - will match functions that have in it These are the only wild cards which are supported. * will not work. # echo hrtimer_* > /debug/tracing/set_ftrace_filter Produces: # tracer: ftrace # # TASK-PID CPU# TIMESTAMP FUNCTION # | | | | | bash-4003 [00] 1480.611794: hrtimer_init <-copy_process bash-4003 [00] 1480.611941: hrtimer_start <-hrtick_set bash-4003 [00] 1480.611956: hrtimer_cancel <-hrtick_clear bash-4003 [00] 1480.611956: hrtimer_try_to_cancel <-hrtimer_cancel -0 [00] 1480.612019: hrtimer_get_next_event <-get_next_timer_interrupt -0 [00] 1480.612025: hrtimer_get_next_event <-get_next_timer_interrupt -0 [00] 1480.612032: hrtimer_get_next_event <-get_next_timer_interrupt -0 [00] 1480.612037: hrtimer_get_next_event <-get_next_timer_interrupt -0 [00] 1480.612382: hrtimer_get_next_event <-get_next_timer_interrupt Notice that we lost the sys_nanosleep. # cat /debug/tracing/set_ftrace_filter hrtimer_run_queues hrtimer_run_pending hrtimer_init hrtimer_cancel hrtimer_try_to_cancel hrtimer_forward hrtimer_start hrtimer_reprogram hrtimer_force_reprogram hrtimer_get_next_event hrtimer_interrupt hrtimer_nanosleep hrtimer_wakeup hrtimer_get_remaining hrtimer_get_res hrtimer_init_sleeper This is because the '>' and '>>' act just like they do in bash. To rewrite the filters, use '>' To append to the filters, use '>>' To clear out a filter so that all functions will be recorded again: # echo > /debug/tracing/set_ftrace_filter # cat /debug/tracing/set_ftrace_filter # Again, now we want to append. # echo sys_nanosleep > /debug/tracing/set_ftrace_filter # cat /debug/tracing/set_ftrace_filter sys_nanosleep # echo hrtimer_* >> /debug/tracing/set_ftrace_filter # cat /debug/tracing/set_ftrace_filter hrtimer_run_queues hrtimer_run_pending hrtimer_init hrtimer_cancel hrtimer_try_to_cancel hrtimer_forward hrtimer_start hrtimer_reprogram hrtimer_force_reprogram hrtimer_get_next_event hrtimer_interrupt sys_nanosleep hrtimer_nanosleep hrtimer_wakeup hrtimer_get_remaining hrtimer_get_res hrtimer_init_sleeper The set_ftrace_notrace prevents those functions from being traced. # echo '*preempt*' '*lock*' > /debug/tracing/set_ftrace_notrace Produces: # tracer: ftrace # # TASK-PID CPU# TIMESTAMP FUNCTION # | | | | | bash-4043 [01] 115.281644: finish_task_switch <-schedule bash-4043 [01] 115.281645: hrtick_set <-schedule bash-4043 [01] 115.281645: hrtick_clear <-hrtick_set bash-4043 [01] 115.281646: wait_for_completion <-__stop_machine_run bash-4043 [01] 115.281647: wait_for_common <-wait_for_completion bash-4043 [01] 115.281647: kthread_stop <-stop_machine_run bash-4043 [01] 115.281648: init_waitqueue_head <-kthread_stop bash-4043 [01] 115.281648: wake_up_process <-kthread_stop bash-4043 [01] 115.281649: try_to_wake_up <-wake_up_process We can see that there's no more lock or preempt tracing. trace_pipe ---------- The trace_pipe outputs the same content as the trace file, but the effect on the tracing is different. Every read from trace_pipe is consumed. This means that subsequent reads will be different. The trace is live. # echo function > /debug/tracing/current_tracer # cat /debug/tracing/trace_pipe > /tmp/trace.out & [1] 4153 # echo 1 > /debug/tracing/tracing_enabled # usleep 1 # echo 0 > /debug/tracing/tracing_enabled # cat /debug/tracing/trace # tracer: function # # TASK-PID CPU# TIMESTAMP FUNCTION # | | | | | # # cat /tmp/trace.out bash-4043 [00] 41.267106: finish_task_switch <-schedule bash-4043 [00] 41.267106: hrtick_set <-schedule bash-4043 [00] 41.267107: hrtick_clear <-hrtick_set bash-4043 [00] 41.267108: wait_for_completion <-__stop_machine_run bash-4043 [00] 41.267108: wait_for_common <-wait_for_completion bash-4043 [00] 41.267109: kthread_stop <-stop_machine_run bash-4043 [00] 41.267109: init_waitqueue_head <-kthread_stop bash-4043 [00] 41.267110: wake_up_process <-kthread_stop bash-4043 [00] 41.267110: try_to_wake_up <-wake_up_process bash-4043 [00] 41.267111: select_task_rq_rt <-try_to_wake_up Note, reading the trace_pipe file will block until more input is added. By changing the tracer, trace_pipe will issue an EOF. We needed to set the function tracer _before_ we "cat" the trace_pipe file. trace entries ------------- Having too much or not enough data can be troublesome in diagnosing an issue in the kernel. The file buffer_size_kb is used to modify the size of the internal trace buffers. The number listed is the number of entries that can be recorded per CPU. To know the full size, multiply the number of possible CPUS with the number of entries. # cat /debug/tracing/buffer_size_kb 1408 (units kilobytes) Note, to modify this, you must have tracing completely disabled. To do that, echo "nop" into the current_tracer. If the current_tracer is not set to "nop", an EINVAL error will be returned. # echo nop > /debug/tracing/current_tracer # echo 10000 > /debug/tracing/buffer_size_kb # cat /debug/tracing/buffer_size_kb 10000 (units kilobytes) The number of pages which will be allocated is limited to a percentage of available memory. Allocating too much will produce an error. # echo 1000000000000 > /debug/tracing/buffer_size_kb -bash: echo: write error: Cannot allocate memory # cat /debug/tracing/buffer_size_kb 85